Chronic alcoholism is well known to affect cerebral function. Nevertheless, the nosology and pathophysiology of alcohol-related brain impairment syndromes remain confusing and incompletely understood. Classifications of such syndromes have characteristically distinguished Korsakoff's syndrome (KS), described as an amnestic disorder consequent to diencephalic pathology, from alcoholic dementia (AlcD), a global disturbance in cognition which is presumably due to diffuse cortical damage. However, the existence of AlcD as a distinct syndrome is controversial, and the overlap between KS and AlcD has become increasingly apparent since KS patients show evidence of cortical impairment and AlcD patients show evidence of subcortical damage. This project will apply the technique of positron emission tomography (PET) combined with careful clinical and neuropsychological characterization of patient groups, to the study of alcohol-related brain impairment syndromes. The PET600 tomograph (2.6 mm resolution) with the glucose metabolic tracer 2-[18F]-fluoro-2-deoxyglucose (FDG) will be used to quantitate regional cerebral metabolic rates for glucose (rCMRglc), thus providing a measurement of regional brain function. The objective of the project is the definition of the brain regions which show impaired function in different alcoholic neurobehavioral syndromes in order to better classify these syndromes as well as understand the mechanisms underlying memory impairment in chronic alcoholism. The study will include subjects from four groups: non-alcoholic controls, non-demented alcoholics, patients with AlcD, and patients with KS. All subjects will undergo a careful medical and neurological screening examination, and will have a battery of cognitive tests designed to characterize and quantitate intellectual deficits. Evaluation of anterograde and retrograde memory, temporal discrimination, and release from proactive interference will be an important part of the test battery. All subjects will have PET-FDG studies as well as magnetic resonance imaging of brain structure. A particular feature of the study is the determination of rate constants for glucose transport and phosphorylation in a subset of patients from each group. The core hypothesis of this project is the expectation that the relative metabolic impairment of cortical and diencephalic brain regions will define the clinical presentation of the alcoholic neurobehavioral syndromes and will be related to the specific character of memory disorders. This work will have important consequences for classification of the alcoholic neurobehavioral syndromes which will allow further investigations in the field to proceed on a firmer footing. Characterization of the metabolic bases of these behavioral deficits will also provide important insights into the neural basis of memory function in humans.